Alzheimer's disease (AD) is the most common cause of dementia after the age of 60. The pathological hallmarks of AD include deposition of amyloid [unreadable]-peptide (A[unreadable]) in neuritic plaques and cerebral blood vessels, neurofibrillary tangles, and loss of neurons. Increasing evidence supports the notion that A[unreadable] and its precursor play important roles in the pathogenesis of AD. Immunization of mouse models of AD with synthetic A[unreadable] reduces A[unreadable] deposits and attenuates their memory and learning deficits. Recent clinical trials, however, were halted due to brain inflammation, presumably induced by T-cell mediated and/or Fc-mediated immune responses. Peripheral administration of antibodies against A[unreadable] also induced clearance of preexisting amyloid plaques in AD mouse models, indicating that an active T-cell-mediated immune response is unnecessary. Topical application of the F(ab')2 without Fc of antibodies against A[unreadable] led to clearance of amyloid deposits in an AD mouse model, indicating that non-Fc-mediated mechanisms are involved in the clearance. Although these passive immunization modalities may be effective in treating AD patients without inducing side effects such as inflammatory responses, such modalities suffer from repeated administrations of antibodies, leading to a large financial and physical burden to AD patients. We hypothesize that single chain antibodies (scFvs) against A[unreadable] are effective and safe in treating AD mouse models. We propose to optimize the efficacy of anti-A[unreadable] scFv delivery in reducing cerebral A[unreadable] load and improving behavioral deficits using three deferent serotypes of recombinant adeno-associated virus (rAAV) vectors encoding anti-A[unreadable] scFv. This study will serve as a proof of principle to demonstrate if this gene therapy modality can deliver anti-A[unreadable] scFvs to many brain regions to reduce A[unreadable] load and improve learning and memory deficits in AD model mice. The following specific aims will be carried out in order to optimize prophylactic and therapeutic delivery of anti-A[unreadable] scFv. In Aim 1, we will evaluate expression and cytotoxicity of scFv delivery by a single intracerebroventricular injection of rAAV1, rAAV8, or AAV9 encoding anti-A[unreadable] scFv in C57BL/6 mice to optimize intracranial delivery of anti-A[unreadable] scFv. In Aim 2, we will evaluate prophylactic and therapeutic effects of intracerebroventricular injections of optimized rAAV encoding anti-A[unreadable] scFv in an AD mouse model. We utilize neuropathological, biochemical, immunological, and behavioral analyses to determine the prophylactic and therapeutic effects and safety of the modality. The long-term goal is to establish the logical basis for developing safe and effective delivery systems of anti-A[unreadable] antibody for AD. Project De isease (AD) is the most common cause of dementia in the elderly. To date, however, no satisfactory treatments are available for AD. This study will serve as a proof of principle to demonstrate if our novel immune gene therapy modality for therapeutic antibody delivery is effective in treating an animal model of AD. Public Hea [unreadable] [unreadable] [unreadable]